>Found part of it --

Found more of it...

The default.wwwtpl file was also polluted with 0xB6 and 0x96 
characters. I also found that the default.mailtpl file has to be 
cleaned of the ^M characters, but the default.wwwtpl file could not be.

At this point, all the java and command kreft are clean and 
everything looks much better -- the management pages all work and the 
main archive page works.

However:

The individual lists' archives cannot be accessed; first, because of 
a problem in the archive page for each list and, second, because of a 
broken template for the list archives themselves.

The individual list archive pages are all created OK except for one 
line at the top:

<META HTTP-EQUIV="Refresh" CONTENT="0;URL=http://URL/cgi-bin/wa?A0=LISTNAME">

That causes the browser to reload the archive using the wa 
executable, which then just loads a shell of a page with no detail at 
all. Here's what every page sees when loaded that way (quoted so it 
will show in all clients):

><br><br> <table width="100%"  cellpadding="0" cellspacing="0" border="0">
><tr> <td align="left"><p><a
>href="{WWW_ARCHIVE_CGI}?INDEX">{NODE}</a></p></td>
><td   align="right"><p>    <a   href="http://www.lsoft.com/catalist.html"
>target="_blank"><img
>src="http://{NODE}/archives/images/b-catalist.gif"
>alt="CataList  Email  List  Search" title="CataList  Email  List  Search"
>border="0"></a> <a
>href="http://www.lsoft.com/products/listserv-powered.asp"
>target="_blank"><img
>src="http://{NODE}/archives/images/b-lpowered.gif"
>alt="Powered by  the LISTSERV Email  List Manager" title="Powered  by the
>LISTSERV Email  List Manager"  border="0"></a> </p> </td>  </tr> </table>
></td></tr></table> </body> </html>

This also shows a deleterious change since it now uses the NODE name 
for the URL. Bad idea. Now none of the images in that default template work.

The search page does work and you can access the individual postings 
from the search results.


-- DCP